world journal of engineering research and technology · deepika et al. world journal of engineering...
TRANSCRIPT
![Page 1: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/1.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
58
REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD
SYSTEM
Deepika Saxena1*, Dr. R.K. Chauhan
2 and Shilpi Saxena
3
1Assisstant Professor, Department of Computer Science and Applications, Kurukshetra
University Kurukshetra, India.
2Professor, Department of Computer Science and Applications, Kurukshetra University
Kurukshetra, India.
3Assisstant Professor, Dpartment of Computer Science and Applications, Graphic Era Hill
University, Dehradun, India.
Article Received on 25/09/2016 Article Revised on 15/10/2016 Article Accepted on 07/11/2016
ABSTRACT
Cloud Computing has extremely surpassed the whole world of
Information Technology. Basically, cloud is a cluster of distributed and
interlinked servers providing on-demand services to customers.
Broadly, it offers software-as-a-service(SAAS), platform-as-a-
service(PAAS), infrastructure-as-a-service(IAAS). Here we are
focusing on IAAS cloud system which offers computational resources
to remote customers in the form of leases. Here we are defining real-time or online optimized
scheduling of requests of various resources arriving simultaneously at data center of IAAS
cloud service provider. Being more practical, our algorithm is providing best resource
utilization and better results in terms of execution time as compared to DFPT algorithm for
task scheduling in cloud computing which do not considers dependency between
tasks(requests). Our algorithm proposes dynamic task allocation on IAAS clouds and
resource scheduling by utilizing the updated status of various Virtual machines available at
real time. We have simulated this experiment using CloudSim toolkit. Surely, there is very
beneficial improvement in results as compared to default FCFS scheduling and other
available scheduling algorithms.
KEYWORDS: Cloud, CloudSim, Dependency, Resource allocation, Task, Virtual Machine.
wjert, 2016, Vol. 2, Issue 6, 58 -70.
World Journal of Engineering Research and Technology
WJERT
www.wjert.org
ISSN 2454-695X Original Article
SJIF Impact Factor: 3.419
*Corresponding Author
Dr. Deepika Saxena
Assisstant Professor,
Department of Computer
Science and Applications,
Kurukshetra University
Kurukshetra, India.
![Page 2: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/2.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
59
INTRODUCTION
In this new digital world of Information Technology and Advancement, the cloud computing
has broaden the world of internet to all most everyone in this world. It has transformed the
complex internet based services to the simplest laymen requirement of computation. The term
``Cloud`` means a cluster of interconnected distributed datacenters which are expanded
worldwide for example, Google, Amazon etc. At each datacenter, there are several computer
systems acting as servers and providing enormous variety of services to the customers.
Briefly describing, cloud computing is a model for enabling convenient, on-demand network
access to shared pool of computing resources for example, networks, servers, storage,
applications and several other services that can be rapidly provisioned and released with
minimal management effort or cloud service provider interaction.[4]
The basic idea of cloud
computing is based on a very fundamental principal of reusability of IT capabilities. In this
paper, specifically considering IaaS clouds i.e. Infrastructure as a Service which represents
lowest level of cloud service providing resources and services on pay-as-per use to its
customers. The ultimate goal of this Cloud System is to instantly provide resources whenever
customer requires it on their laptops, PCs, mobile phones etc wherever possible. In an IaaS
model of cloud computing, Cloud Service Provider provides hardware, software, servers,
storage and other infrastructure components to its customers. IaaS clouds also provides
various applications to its users and handle task including system maintenance, backup and
resiliency planning. IaaS customers pay on a per-use basis typically by hour, week or month.
Some Cloud Service Provider also charge customers based on amount of storage space used
or VM space used.[5]
In recent trends of IT world, cloud computing is emerging with growing
popularity of its capability to provide unlimited VM space, storage space and other resources
with extreme flexibility so that no adjustment is required at customer end. But practically,
there is no datacenter which has unlimited capacity. Infact, to meet real time significant
demand of resources, Cloud Service Provider has to make several arrangements and
adjustments at their datacenters to concurrently schedule the thousand of requests and
demand of resources arriving parallel at datacenter. In case of overflow of workload at
datacenter, it can be shared among different datacenters or workload can be share in between
public and private clouds with applying extreme security checks.[2]
So, this becomes a very
challenging issue to schedule several requests simultaneously at datacenter satisfying all the
customers, workload balancing, reduction of energy consumption, faster execution of
requests at minimum cost etc. Clearly, default FCFS scheduling cannot provide satisfactory
results in all the above mentioned respect. Therefore, there is rigorous demand of an
![Page 3: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/3.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
60
optimized real-time scheduling technique. Though, many researchers have proposed different
algorithms for scheduling in cloud computing which are providing good solutions in one or
other respect, but none of them is able to completely satisfy all the different criteria of
scheduling in cloud system. Broadly, in IaaS cloud system there is requirement of a
scheduling technique which can satisfy all the different terms and conditions of real-time
cloud computation. The scheduling technique should consider dependency and non-
dependency among tasks or requests, bandwidth utility, data transfer rate, data transfer cost,
response-time of VMs, workload balancing, resource sharing, CPU utilization, faster
execution-time at minimum cost, how many VMs to be set up at one host computer system
and many more aspects like that. Here, in our proposed work we have surely covered all
these aspects and designed a satisfactory algorithm that is providing improved results in all
respect.
LITRETURE SURVEY
Combined resource provisioning and scheduling strategy for executing scientific workflows
on IaaS Clouds was presented which aims to minimize overall execution cost while meeting a
user defined deadline.[6]
This paper provides an optimized solution to scheduling problem of
scientific workflows in cloud system. But it is considering single workflow execution at a
time and not the multiple workflows. Various energy efficient strategies were studied which
are based on approaches that specific plug-ins and energy control centers for networked
large-scale hardware and software can be implemented and that they can reduce software and
hardware related energy cost, improve load-balancing, reduce energy consumption due to
communications, save CO2 emissions etc.[1]
This paper has considered energy saving aspects
only and not noticed the execution time, execution cost and dependency between tasks in
cloud computing system. In case of significant client demands, it is necessary to share
workload among multiple datacenters, this would provide cheaper resources and would
enlarge the capacity of resource pool.[2]
This paper is very flexible and is efficient in
providing improved results for task scheduling in IaaS clouds. Also, they have designed two
online dynamic scheduling algorithms i.e. DCLS and DCMMS for resource allocation
mechanism. In DFPT algorithm[3]
weighted fair priority queue is applied to schedule task
according to deadline based criteria and minimum cost-based criteria. Though results are
optimized and improved in this algorithm but it has limitations like it is not suitable
dependent tasks, energy saving and load balancing criteria are not defined.
![Page 4: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/4.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
61
PROPOSED MODEL CONCEPTS
Our proposed model includes the following concepts: Here we assume that several requests
are arriving at datacenter simultaneously, these requests are termed as tasks in cloud
computing. In reality, most of the request shows dependencies upon one another, so here we
have schedule the tasks according to their dependencies, also complexity of computation
increases along with increasing dependencies. The set of tasks which includes dependencies
represents ``workflows``.
Following steps are followed in our proposed model:
1. Preparation of list of workflows (or set of related tasks) submitted by several users at
data center.
2. Independent Tasks: All independent task at same level can execute in parallel separately
on selected VMs.
3. Dependent Tasks: All tasks which have output dependency on another task, their
transfer time and transfer cost is calculated as follows:
Transfer Time = Data output / bandwidth utilized for data transfer
Transfer Cost = Data output / cost of bandwidth used.
4. Bandwidth at same datacenter will cost same for data transfer, but if data is transferred to
some external datacenter it will cost different and it will be high.
5. Whenever a task is selected from the workflow for execution, it is mandatory to execute
all its predecessor task first to resolve the dependency between tasks.
6. Execution Time and Execution Cost of a single task is calculated as follows:
Execution Time = Task length / Processing power of selected VM
Execution Cost = Cost of selected VM * Task length
Here task length is number of instructions of a task to be executed.
7. Finally, Total Processing time and cost of each workflow is calculated as:
Total Processing Time = sum of execution time of each task in the task set + sum of all
transfer time related to a task set.
Total Processing Time = sum of execution cost of each task in the task set + sum of
transfer cost of the task set.
8. Resource allocation is done using greedy approach of modified Cloud min-min
algorithm, which selects suitable resource i.e. VM that execute the selected task at
minimum cost and minimum time. Separate list of resource is maintained which keep
track of the status of the VM by continuous VM monitoring at each resource allocation.
![Page 5: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/5.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
62
9. Workload balancing is done as follows: First of all, workload is calculated which
depends upon the complexity and amount of computation involved in the task set. The
complexity can be measured in terms of dependencies and communication among the
tasks, amount of expected execution time based upon number of instructions in the task ,
memory space required for execution of task etc. If the calculated workload is higher than
the capacity of single server, then workload is distributed among other servers at same
datacenter, or if required among servers at different datacenter to balance the workload.
Hence, workload is shared among distributed servers at same or different data center.
Figure 1: Applications represented as DAG.
Figure 2: List of Applications.
Figure 1 and 2 are showing applications as submitted by the user, each application is a
complex, so it is partitioned into set of task which are interconnected to each other in various
respects, to show their relationship we have taken each application as DAG direct acyclic
![Page 6: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/6.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
63
graph which clearly shows the dependency among tasks as possess by the real request at the
cloud server at datacenter.
METHODOLOGY
User submits applications to be executed on cloud server at datacenter as shown in figure 3.
Each application is first partitioned into tasks in the form of DAG (Direct Acyclic Graph)
based on dependency among tasks. Then these DAGs are placed in the list based on priorities
(like deadline, cost, complexity etc) of execution using algorithm 1.
These priorities based list is then submitted at different cloud servers where they are
scheduled and assigned to different VMs or resources using real-time resource allocation
procedure as described in algorithm 2.
Figure 3: When an application is submitted to the cloud system, it is partitioned,
assigned, scheduled and executed in the cloud system.[2]
ALGORITHM 1: Preparation of task list based on priorities.
Requirement – A DAG, expected execution time of each task in DAG
1. Deadline of every task is randomly generated.
2. Initially, all tasks are in list U.
3. Take task one by one from U and push current task node into stack S in decreasing order
of their deadline.
4. While stack S is not empty do
5. If top(S) has unstacked immediate predecessor task then
6. Push immediate predecessor task node with least deadline into stack S.
7. Else push top of stack S into list P.
![Page 7: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/7.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
64
8. Pop top(S).
9. EndIf.
10. EndWhile.
Result- A list of task of P is prepared based on priorities.
ALGORITHM2: Real-time resource allocation Cloud min-min scheduling(RCMMS).
Requirement – Priority based list of task P, n different VMs, execution-time matrix and data
transfer-time matrix.
1. Generate mappable task set P using above algorithm 1.
2. While tasks not assigned do
3. T = top(P).
4. Get resource status information from all other DataCenterBroker(DCB) or manager
servers.
5. Find resource or VM with minimum (VMmin) turnaround time (= waiting time + response
time + execution time).
6. Assign task T to VMmin.
7. Remove T from P.
8. EndWhile.
Result – Selected task T is assigned for execution to VM with earliest finish time.
Performance Evaluation
In this section we have presented the experiments conducted in order to evaluate the
performance of the proposed work. We utilized the CloudSim 3.1 tool to simulate the cloud
environment and task sets are generated randomly using rand() function. The proposed work
is compared with sequential, DFPT, EWSA algorithm for cloudlet scheduling. The
configuration of the datacenter are as follows:
Table 1 shows the configuration of host and table 2 shows the configuration of VMs used in
this proposed work. Here, we have used:
Number of processing element – 1
Number of host on each processor – 8 on each processor.
Table 1: Configuration of host.
RAM (MB) 2080
Processing Power (MB) 220000
VM Scheduling Time-shared
![Page 8: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/8.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
65
Table 2: Configuration of VMs.
VM RAM Processing
Power(MIPS)
Processing
Element
VM1 5024 22000 1
VM2 1048 11000 1
VM3 3308 22000 1
VM4 4604 32000 1
VM5 8028 55000 1
VM6 4000 41000 1
VM7 6024 44000 1
VM8 7028 62000 1
Table 3: System Configuration.
Processor Pentium Dual-core CPU [email protected]
RAM 3.00GB
System Types 64 bit operating system
Operating System Windows 7
Performance with respect to time: The experimental results shows the remarkable
improvement in time over sequential, DFPT and EWSA algorithms.
Table 4: Performance of proposed methodology with respect to time.
No. of task set FCFS DFPT EWSA PRPOSED
METHODOLOGY
10 803.6499 366.2881 359.4416 280.6669
20 3202.9668 1188.6538 1146.1112 660.5355
30 5564.0266 3766.8554 2881.6667 1044.6681
40 7004.3999 5044.6669 3331.1161 1846.4401
50 8963.3897 6887.4331 5842.556 3566.6664
60 10468.26666 8113.9988 7046.1134 5442.1448
Figure 4: Graph showing comparision of execution time of proposed work RMCC with
existing algorithms.
![Page 9: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/9.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
66
Performance with respect to cost: The experimental results shows the immense
improvement in cost over sequential, DFPT and EWSA algorithms.
Table 5: Comparision of Cost of Execution of Proposed methodology V/s existing
algorithms.
No. of task set FCFS DFPT EWSA PROPOSED
METHODOLOGY
10 363.44 200.448 197.576 160.001
20 539.61 369.866 311.667 271.904
30 944.68 677.007 661.112 511.008
40 1248.56 721.481 624.081 441.169
50 1554.81 1098.224 976.116 771.816
60 1768.66 1544.616 1496.672 981.197
Figure 5: Graph showing comparision of cost of execution of proposed RMCC method
with existing algorithms.
Table 4 and Table 5 are showing performance evaluation of various number of task set
against execution time and execution cost. Here, a task set can have one task, two tasks or up
to five tasks having interdependency. A task set is generated at run time and dependency is
also decided at execution time randomly. So we have calculated, the results of various
execution time of task set as the arithmetic mean of execution time obtained on running the
task set twenty times. Each time result may somewhat vary depending upon dependency and
number of task in task set.
CONCLUSION AND FUTURE WORK
IAAS cloud computing offers immense research work to do like data security, virtualization
license management etc. But the scheduling is the topic of research in this field. Many well
![Page 10: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/10.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
67
defined algorithms are proposed by many researchers, but none of them is suitable in all the
way. In contrast, our proposed work is more dynamic and pragmatic that meets all the
demands of perfect scheduling in IAAS. We have proposed RCMMS and workload balancing
concepts and resolved dependency among task sets while minimizing execution time and
execution cost. We have provided complete scenario of real time IAAS cloud scheduling and
implemented it using CloudSim. The outcome results are showing tremendous improvement
as compared to FCFS, DFPT[3]
and EWSA[6]
cloudlet scheduling algorithms.
In future more sophisticated and real algorithms can be research and work out for IAAS
cloud system including better techniques for workload balancing, energy-saving techniques
and considering contention problem at datacenters.
REFRENCES
1. Andreas Berl, Erol Gelenbe, Marco Di Girolamo, Giovanni Giuliani, Hermann De Meer,
Minh Quan Dang and Kostas Pentikou sis, “Energy-Efficient Cloud Computing”,
Advance Access publication on August 19, 2009.
2. Jiayin Li, Meikang Qiu, Zhong Ming, Gang Quan, Xiao Qin, Zonghua Gu, “Online
optimization for scheduling preemptable tasks on IaaS cloud system”, J.Parallel
Distributed Computing journal, 2012; 72: 666-677.
3. Deepika Saxena, Dr. R.K. Chauhan, Dr. Ramesh Kait,”Dynamic Fair Priority
Optimization Task Scheduling Algorithm in Cloud Computing: Concepts and
Implementation”, I.J. Computer Network and Information Security”, 2016; 2: 41-48.
4. P.Mell and T.Grance, “The NIST definition of cloud computing”, National Institute of
Standard and Technology, 2009; 53.
5. Infrastructure as a Service,
http://searchcloudcomputing.techtarget.com/definition/Infrastructure-as-a-service.
6. Shilpi Saxena, Deepika Saxena,”Enriched Workflow Scheduling Algorithm”, IEEE
International Conference on Computing, Communication and Security(ICCCS), 2015
10.1109/CCCS.2015.7374202
7. H.J.M.Y. Chi, H. Hacigumus, Performance evaluation of scheduling algorithmsfor
database services with soft and hard SLAs, in: International Workshop onData Intensive
Computing in the Clouds, 2011; 1–10.
8. J. Dean, S. Ghemawat, MapReduce: simplified data processing on large
clusters,Communications of the ACM, 2008; 1: 107–113.
![Page 11: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/11.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
68
9. A. Dogan, F. Ozguner, Matching and scheduling algorithms for minimizingexecution
time and failure probability of applications in heterogeneous computing, IEEE
Transactions on Parallel and Distributed Systems, 2002; 13(3): 308–323.
10. W. Emeneker, D. Stanzione, Efficient virtual machine caching in dynamic virtual
clusters, in: SRMPDS Workshop of International Conference on Parallel and Distributed
Systems, Hsinchu, Taiwan, 2007.
11. N. Fallenbeck, H. Picht, M. Smith, B. Freisleben, Xen and the art of cluster scheduling,
in: Proceedings of the 2nd International Workshop on Virtualization Technology in
Distributed Computing, Tampa, Florida, USA, 2006; 4.
12. T. Forell, D. Milojicic, V. Talwar, Cloud management challenges and opportunities, in:
IEEE International Symposium on Parallel and Distributed, 2011; 881–889.
13. G. Garzoglio, T. Levshina, P. Mhashilkar, S. Timm, ReSS: a resource selection service
for the open science grid, in: International Symposium on Grid Computing, 2007; 1–10.
14. C. Germain-Renaud, O. Rana, The convergence of clouds, grids, and autonomics, IEEE
Internet Computing 2009; 13(6): 9.
15. Apache Hadoop, http://wiki.apache.org/hadoop/.
16. T. Hagras, J. Janecek, A high performance, low complexity algorithm for compile-time
task scheduling in heterogeneous systems, Parallel Computing, 2005; 31(7): 653–670.
17. H. Huang, L. Wang, P&p: a combined push–pull model for resource monitoring in cloud
computing environment, in: IEEE International Conference on Cloud Computing, Miami,
Florida, USA, 2010; 260–266.
18. O.H. Ibarra, C.E. Kim, Heuristic algorithms for scheduling independent tasks on
nonidentical processors, Journal of the ACM., 1977; 24(2): 280–289.
19. D. Irwin, J. Chase, L. Grit, A. Yumerefendi, D. Becker, K.G. Yocum, Sharing networked
resources with brokered leases, in: USENIX Annual Technical Conference, Boston,
Massachusetts, USA, 2006.
20. W. Jiang, Y. Turner, J. Tourrilhes, M. Schlansker, Controlling traffic ensembles in open
cirrus, https://www.hpl.hp.com/techreports/2011/HPL-2011-129.pdf.
21. K. Keahey, T. Freeman, Contextualization: providing one-click virtual clusters, in: IEEE
International Conference on eScience, Indianapolis, Indiana, USA, 2008.
22. M.A. Kozuch, M.P. Ryan, R. Gass, S.W. Schlosser, D. O’Hallaron, et al. Cloud
management challenges and opportunities, in: Workshop on Automated control for
datacenters and cloud, 2009; 43–48.
![Page 12: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/12.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
69
23. H. Liu, D. Orban, GridBatch: cloud computing for large-scale data-intensive batch
applications, in: IEEE International Symposium on Cluster Computing and the Grid,
2008; 295–305.
24. Open computing facility - ocf, https://computing.llnl.gov/?set=resources&page=OCF-
resource.
25. R. Moren-Vozmediano, R.S. Montero, I.M. Llorente, Elastic management of cluster-
based services in the cloud, in: Workshop on Automated control for datacenters and
cloud, 2009; 19–24.
26. C. Moretti, J. Bulosan, P.J. Thain, D. Flynn, All-pairs: an abstraction for dataintensive
cloud computing, in: IEEE International Symposium on Parallel and Distributed
Processing, 2008; 1–11.
27. A. Pavlo, E. Paulson, A. Rasin, D.J. Ababi, D. DeWitt, S. Madden, M. Stonebraker, A
comparison of approaches to large-scale data analysis, in: SIGMOD, 2009; 1–14.
28. Parallel workloads archive, www.cs.huji.ac.il/labs/parallel/workload/.
29. M. Qiu, M. Guo, M. Liu, C.J. Xue, L.T. Yang, E.H.-M. Sha, Loop scheduling and bank
type assignment for heterogeneous multi-bank memory, Journal of Parallel and
Distributed omputing (JPDC) 2009; 69(6): 546–558.
30. M. Qiu, E.H.-M. Sha, Cost minimization while satisfying hard/soft timing constraints for
heterogeneous embedded systems, ACM Transactions on Design Automation of
Electronic Systems (TODAES) 2009; 14(2): 1–30.
31. B.P. Rimal, E. Choi, I. Lumb, A taxonomy and survey of cloud computing systems, in:
International Joint Conference on INC, IMS and IDC, 2009; 44–51.
32. P. Ruth, P. McGachey, D. Xu, Viocluster: virtualization for dynamic computational
domains, in: Proceedings of the IEEE International Conference on Cluster Computing,
Boston, Massachusetts, USA, 2005; 1–10.
33. P. Ruth, J. Rhee, D. Xu, R. Kennell, S. Goasguen, Autonomic live adaptation of virtual
computational environments in a multi-domain infrastructure, in: IEEE International
Conference on Autonomic Computing, 2006; 5–14.
34. W. Smith, I. Foster, V. Taylor, Scheduling with advanced reservations, in: IEEE
International Parallel and Distributed Processing Symposium, Cancun, Mexico, 2000;
127–132.
35. B. Sotomayor, K. Keahey, I. Foster, Combining batch execution and leasing using virtual
machines, in: Proceedings of the 17th International Symposium on High Performance
Distributed Computing, Boston, Massachussets, USA, 2008; 87–96.
![Page 13: World Journal of Engineering Research and Technology · Deepika et al. World Journal of Engineering Research and Technology WJERT 58 REAL TIME OPTIMIZATION SCHEDULING ON IAAS CLOUD](https://reader033.vdocuments.us/reader033/viewer/2022050212/5f5e903b67a3be752327b7ce/html5/thumbnails/13.jpg)
www.wjert.org
Deepika et al. World Journal of Engineering Research and Technology
70
36. B. Sotomayor, R. Llorente, I. Foster, Resource leasing and the art of suspending virtual
machines, in: 11th IEEE International Conference on High Performance Computing and
Communications, Seoul, Korea, 2009; 59–68.
37. B. Sotomayor, R. Montero, I. Llorente, I. Foster, Virtual infrastructure management in
private and hybrid clouds, IEEE Internet Computing 2009; 13(5): 14–22.
38. Amazon AWS, http://aws.amazon.com/.
39. GoGrid, http://www.gogrid.com/.
40. RackSpace, http://www.rackspacecloud.com/.
41. Microsoft cloud, http://www.microsoft.com/en-us/cloud/.
42. IBM cloud, http://www.ibm.com/ibm/cloud/.
43. Google apps, http://www.google.com/apps/intl/en/business/index.html.
44. HP cloud, http://www8.hp.com/us/en/solutions/solutionsdetail.html? compURI=tcm:245-
300983\&pageTitle=cloud.
45. Eucalyptus, http://www.eucalyptus.com/.
46. RESERVOIR, www.reservoir-fp7.eu.
47. E. Walker, J. Gardner, V. Litvin, E. Turner, Dynamic virtual clusters in a grid site
manager, in: Proceedings of Challenges of Large Applications in Distributed
Environments, Paris, France, 2006; 95–103.
48. X. Wang, J. Zhang, H. Liao, L. Zha, Dynamic split model of resource utilization in
mapreduce, in: International Workshop on Data Intensive Computing in the Clouds,
2011; 1–10.
49. M. Wilde, M. Hategan, J.M. Wozniak, B. Clifford, D.S. Katz, I. Foster, Swift: a language
for distributed parallel scripting, Parallel Computing 2011; 37(9): 633–652.
50. T. Wood, A. Gerber, K. Ramakrishnan, J. van der Merwe, The case for enterpriseready
virtual private clouds, in: Workshop on Hot Topics in Cloud Computing, San Diego,
Califoria, USA, 2009.
51. S. Zanikolas, R. Sakellariou, A taxonomy of grid monitoring systems, Future Generation
Computer systems, 2005; 1: 163–188.